Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
194 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Realtime Limb Trajectory Optimization for Humanoid Running Through Centroidal Angular Momentum Dynamics (2501.17351v2)

Published 29 Jan 2025 in cs.RO

Abstract: One of the essential aspects of humanoid robot running is determining the limb-swinging trajectories. During the flight phases, where the ground reaction forces are not available for regulation, the limb swinging trajectories are significant for the stability of the next stance phase. Due to the conservation of angular momentum, improper leg and arm swinging results in highly tilted and unsustainable body configurations at the next stance phase landing. In such cases, the robotic system fails to maintain locomotion independent of the stability of the center of mass trajectories. This problem is more apparent for fast and high flight time trajectories. This paper proposes a real-time nonlinear limb trajectory optimization problem for humanoid running. The optimization problem is tested on two different humanoid robot models, and the generated trajectories are verified using a running algorithm for both robots in a simulation environment.

Summary

  • The paper presents a real-time optimization framework that uses centroidal angular momentum dynamics to refine limb trajectories for stable humanoid running.
  • It converts complex nonlinear dynamics into a reduced-size constrained problem, enabling practical real-time performance on different robot models.
  • Simulations on models like Kangaroo and Unitree G1 demonstrate minimized rotational velocities and improved touchdown stability, highlighting broad applicability.

Real-time Limb Trajectory Optimization for Humanoid Running through Centroidal Angular Momentum Dynamics

The paper, authored by Sovukluk, Schuller, Englsberger, and Ott, addresses a pertinent challenge in the domain of humanoid robotics, specifically focusing on optimizing limb trajectory during humanoid running phases. At its core, this work seeks to develop a real-time optimization framework that drives limb trajectories to maintain stability, with a particular emphasis on the conservation of angular momentum during flight phases of humanoid locomotion.

Key Contributions and Methodology

One of the central goals in humanoid locomotion is the refinement of limb trajectories to ensure that the robotic system can maintain an upright posture upon touchdown after a flight phase, devoid of ground reaction forces. The significance of this challenge is elevated in scenarios involving fast humanoid movements with extended flight times, where improper management can result in unsustainable body configurations.

The authors effectively leverage the properties of the centroidal momentum matrix (CMM) to structure their optimization framework. By dissecting the dynamics and distilling the angular and translational components, they provide insightful properties that enable the formulation of a reduced size, nonlinearly constrained optimization problem executable in real-time. This reduction is achieved via targeted simplifications, demonstrating an understanding of which joint movements significantly impact dynamic stability.

Simulation and Results

The proposed methodology was tested and verified in a simulated environment on two humanoid robot models—Kangaroo and Unitree G1. For the Kangaroo model, a simulation running algorithm was executed with a set trajectory (1 m/s forward velocity and 0.31-second flight time). The optimization's primary success is observed in minimizing rotational velocities, attaining a minimal body orientation at touchdown, and ensuring that disturbances do not lead to failure of motion. The simulation results for Unitree G1 mirror these observations while demonstrating the model-agnostic nature of the proposed optimization solution.

Implications and Future Directions

From both practical and theoretical standpoints, this research extends the legged locomotion literature by integrating centroidal angular momentum considerations into trajectory optimization frameworks. By structuring the optimization independent of specific running models, it allows more general application across different humanoid designs and running algorithms. This represents a significant advancement over previous models that heavily relied on predefined or simplistic leg swing patterns.

The potential for future work includes extending this optimization to account for environmental variability and more complex movement patterns involving external perturbations or multi-contact scenarios. Another interesting avenue would be the translational application of these principles to more varied robotic forms or even adaptations for human-driven exoskeletons, marking a boundary-exploration in applied biomechanics and robotics.

In conclusion, the paper offers a comprehensive analysis and solution to the issue of maintaining dynamic stability through effective limb trajectory optimization during the flight phases of humanoid running. The strategies presented facilitate an improved application of nonlinear optimization in real-time settings, bolstering the capabilities of humanoid robots and advancing the field of robotic locomotion.

X Twitter Logo Streamline Icon: https://streamlinehq.com
Youtube Logo Streamline Icon: https://streamlinehq.com